The experience gained with the continuous elimination techniques in patients with acute renal failure, together with the clearance, which, in the case of haemofiltration is independent of molecular size over a wide range of molecular weights, prompt the question as to whether the course of multiple organ failure might not be positively affected by means of an (unspecific) elimination of toxin and mediators via CAVH. Indeed, findings established in the dog and pig in endotoxin shock to point to a possible use of haemofiltration; and the same applies to gram-positive sepsis. However, a positive of haemofiltration has been demonstrated in only some of the models. Although numerous mediators and toxins are known to pass across the membrane, quantitative data on the extent of toxin elimination with respect to the size and turnover rate of the endogenous pool are lacking, as is an estimation of the net effect with simultaneous elimination of mediators and antimediators. Until a relevant advantage of haemofiltration in practice-oriented sepsis models has been confirmed, not only for one-time or brief exposure to toxin or pathogens, as well as in prospective clinical trials, an expansion of the indication spectrum to include cases apart from acute renal failure would appear premature at this time.